In runing betting on football

I thnk some of the trading software out there will do it for you. Not sure how accurate it is though.

I would prefer, if possible, to understand the maths first.
Not to put the horse before the cart and all that stuff.

To help you understand, look at the prices for half time correct score in relation to full time correct score

In what manner exactly ?
Are you saying that the full time prices at the off are related to the HT prices at the off.
And so if a certain score is happening at HT, does that relationship to the now projected FT score at the HT stage, stay the same or change in some manner ?

It's just a projection based on the likelihood of the correct score as the match progresses. Other factors that might affect the price are sendings off and things like that.

I would concur with rocket, however in Pullein's book
''Betting on Football'' he does a piece on 'Correct
Scores', although not in running prices. I would
imagine the calculation to be quite complex, and
computer generated.

Use poisson as a rough approximation?

If you know the implie probability of a team to win then you can also find out the implied probability of them winning by a certain margin.

As Ken says, Kevin Pullein has some look up charts in his book which give a rough guide but these are presumably for matches that haven't yet started and are also from a purely statistical approach.

It can be interesting to look at the amounts matched on each score, for example the next in-play match...

LDU (1.65) v Newels, 02.15 ko, poor liquidity but not unplayable by any means.

Most money is shown as matched on 0-0, 0-1, 3-0, UQ, respectively. But the most money has actually been matched on 3-3, even though the matched amount is currently only £118.

That's just changed slightly as a punter has just backed the 0-0 for around £150 and layed the UQ for 200.

Being able to work out this kind of stuff accurately is how people make money.

Rocket, I don't think that's correct. It's fine if you're looking for an estimate, but you won't get complete accuracy.

Let's say Barcelona are playing a Primera match and Chelsea are playing an EPL match. Both are heavy favourites, priced at 1.50. Without any further information, I would be fairly confident that Barcelona winning 3 goals is more likely than Chelsea winning by three goals. What do you think?

I expect that given only correct score prices (in a liquid market) and no other information, it's possible to work out the half time score implied odds. I've been thinking about doing this, but you would need to build a database of odds I think. Currently I use a different method, but it's tricky getting matched pre-off due to the increasing level of automation. An interesting thing I've noticed is the significant difference in the chance of second half outcomes versus first half outcomes.

Poisson breaks down when goal is expectancy is less than 1.

Hence, a 0-0 "gubs" the stats.

However, it remains a useful tool for populating the prices of a 17-runner outcome C/S market.

Everything else of interest - match odds/overs-unders/Asian H'Caps - can be priced off of C/S.

Learn to use Poisson/compound poisson to price C/S; and you will get there.

No shortcuts though I am afraid.

Baggers Joined: 18 Dec 02
Replies: 7238 13 Oct 10 23:41   


To help you understand, look at the prices for half time correct score in relation to full time correct score 

outstanding post.

Taking the Aston Villa match on Saturday, I currently have

1-0 @  16
2-0 @  32
2-1 @  19.5
3-1 @  55


Which is pretty close to the market at the moment.

What do you get, Investor?

I agree that it is only an estimate.

And

0-1 @ 7.2
0-2 @ 10
1-2 @ 9.2
1-3 @ 20

The stand out incorrect one there is the 0-1

I price up all the main football markets using non-homogenous Poisson distributions. Knowing precisely the respect in which the distributions are non-homogenous is the hard bit.

But 0-1 is can be underestimated when a team is favs

Prices are based a lot on sentiment and flow of play,therefore prices can and DO change ,bear this in mind.

Also, study the brain of a footballer, how does his mind think ?

If a strong side goes 0-2 down at half ,how does this affect the likelihood of say a 3rd goal ? 


Study this and you will be OK [;)]

There's also a lot of maths involved in correct score betting , the ability to do this in your head many times very quickly  without the aid of charts and graphs is PARAMOUNT.

The brain thinks ,a computer doesn't.

There i am giving a sermon ,and then a few minutes later i have my worst loss to date on correct score market 

Soccer / Buyuksehir v Kasimpasa : Correct Score     11-Nov-10 11:30     11-Nov-10 13:24    -398.62

There's also a lot of maths involved in correct score betting , the ability to do this in your head many times very quickly  without the aid of charts and graphs is PARAMOUNT.

The brain thinks ,a computer doesn't.


My computers do all this thinking for me!

Thinking computers havent been invented yet , are you from a differrent galaxy ?

Let's say Barcelona are playing a Primera match and Chelsea are playing an EPL match. Both are heavy favourites, priced at 1.50. Without any further information, I would be fairly confident that Barcelona winning 3 goals is more likely than Chelsea winning by three goals. What do you think?


Over the course of thousands of bets it will verge with the line but you are right, on a game to game basis there will be points well away from the mean

some teams are priced up by name